[Lustre-discuss] MDT fails to mount after lustre upgrade

[Anjana Kar]

Would it make sense to run
"tunefs.lustre --mgs --writeconf --mgs --mdt /dev/sda /dev/sdb" ?

The original mkfs command used to create the MDT was
mkfs.lustre --reformat --fsname=iconfs --mgs --mdt --backfstype=zfs 
--device-size 131072 \
--index=0 lustre-mgs-mdt/mgsmdt0 mirror /dev/sda /dev/sdb

Not sure if both device names should be included or the zpool name.

The latest 2.5.x version also fails to mount MDT with similar messages, 
though
zpool seems intact... is there anyway to get MDT to mount?

[root at icon0 kar]# more /proc/fs/lustre/version
lustre: 2.5.58
kernel: patchless_client
build: 2.5.58-g5565877-PRISTINE-2.6.32-431.11.2.el6.netboot

[root at icon0 kar]# /sbin/service lustre start mgsmdt
Mounting lustre-mgs-mdt/mgsmdt0 on /mnt/lustre/local/mgsmdt
mount.lustre: mount lustre-mgs-mdt/mgsmdt0 at /mnt/lustre/local/mgsmdt 
failed: File exists

[root at icon0 kar]# zpool list
NAME             SIZE  ALLOC   FREE    CAP  DEDUP  HEALTH ALTROOT
lustre-mgs-mdt  74.5G  47.2G  27.3G    63%  1.00x ONLINE  -

Corresponding console messages:
2014-04-24T10:36:53.736967-04:00 icon0.psc.edu kernel: Lustre: Lustre: 
Build Version: 2.5.58-g5565877-PRISTINE-2.6.32-431.11.2.el6.netboot
2014-04-24T10:36:53.986362-04:00 icon0.psc.edu kernel: LNet: Added LNI 
10.10.101.160 at o2ib [8/256/0/180]
2014-04-24T10:36:54.005177-04:00 icon0.psc.edu kernel: LNet: Added LNI 
128.182.75.160 at tcp10 [8/256/0/180]
2014-04-24T10:36:54.009556-04:00 icon0.psc.edu kernel: LNet: Accept 
secure, port 988
2014-04-24T10:36:57.428805-04:00 icon0.psc.edu kernel: LustreError: 
11-0: iconfs-MDT0000-lwp-MDT0000: Communicating with 0 at lo, operation 
mds_connect failed with -11.
2014-04-24T10:36:58.600958-04:00 icon0.psc.edu kernel: LustreError: 
11888:0:(mdd_device.c:1050:mdd_prepare()) iconfs-MDD0000: failed to 
initialize lfsck: rc = -17
2014-04-24T10:36:58.600978-04:00 icon0.psc.edu kernel: LustreError: 
11888:0:(obd_mount_server.c:1776:server_fill_super()) Unable to start 
targets: -17
2014-04-24T10:36:58.614082-04:00 icon0.psc.edu kernel: Lustre: Failing 
over iconfs-MDT0000
2014-04-24T10:37:04.885917-04:00 icon0.psc.edu kernel: Lustre: 
11888:0:(client.c:1912:ptlrpc_expire_one_request()) @@@ Request sent has 
timed out for slow reply: [sent 1398350218/real 1398350218]  
req at ffff8802fca1a000 x1466276472946828/t0(0) 
o251->MGC10.10.101.160 at o2ib@0 at lo:26/25 lens 224/224 e 0 to 1 dl 
1398350224 ref 2 fl Rpc:XN/0/ffffffff rc 0/-1
2014-04-24T10:37:05.418516-04:00 icon0.psc.edu kernel: Lustre: server 
umount iconfs-MDT0000 complete
2014-04-24T10:37:05.418535-04:00 icon0.psc.edu kernel: LustreError: 
11888:0:(obd_mount.c:1338:lustre_fill_super()) Unable to mount  (-17)

TIA,
-Anjana Kar
  Pittsburgh Supercomputing Center
  kar at psc.edu